基于形态特征的玉米种子表面裂纹检测方法研究
首发时间:2008-03-14
摘要:研究采用数字图像处理技术实现对玉米种子表面裂纹的识别和检测。首先选择CCFL(冷阴极荧光灯)设计了图像采集的光照环境,建立了玉米种子图像的采集系统,然后针对玉米种子图像提出了一种基于籽粒形态学特征的表面裂纹检测方法。该方法首先采用水平和垂直边缘检测算子处理得到裂纹、种子边界和噪声等边缘信息;然后通过玉米籽粒的形态特征寻找其尖端位置并使用图像代数运算的方法去除大部分非裂纹信息;最后根据裂纹的长度和位置特征提取得到裂纹,并计算裂纹的绝对长度和相对长度。对农大4967和农大3138两个品种玉米分别选取裂纹粒和无裂纹粒各50粒进行图像识别,试验结果表明:识别准确率分别为94%和90%。基本满足玉米种子表面裂纹检测的精度要求。
For information in English, please click here
Detection of surface cracks of corn kernel based on morphology
Abstract:Surface cracks identification and detection of corn kernel was studied based on digital image processing. CCFL (Cold Cathode Fluorescent Lamps) were chosen to construct the image capturing illumination environment, and a set of image acquisition system of corn kernel was established. Then, a method of surface cracks detection was developed base on morphology of corn kernel. Firstly, binary image including the information of cracks, boundary and noises were picked up with horizontal and vertical Sobel operators. Subsequently, a majority of non-cracks information was eliminated by images subtraction after finding the tip of the corn kernel based on its morphology. Finally, according to the cracks length and position, the cracks were extracted, and the absolute and relative length of cracks were calculated. A detecting experiment was carried out with 50 kernels with cracks and 50 kernels without cracks selected from NongDa-4967 and NongDa-3138 (two novel varieties of corn seed developed by China Agricultural University) respectively. The results indicate that the detecting accuracy were 94% and 90%. It can satisfy the accuracy of surface cracks detection of corn kernels by the method.
Keywords: Computer vision Image processing Corn kernel Surface cracks Morphology
论文图表:
引用
No.1932020670612054****
同行评议
共计0人参与
勘误表
基于形态特征的玉米种子表面裂纹检测方法研究
评论
全部评论0/1000